The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest HTML content (site.html_content) from the user or previous agent steps and incorporate it into generated site files.
Ingestion points: The site.html_content field in the input schema within SKILL.md is the primary entry point for untrusted data.
Boundary markers: The workflow does not specify the use of delimiters or specific instructions to the agent to ignore embedded commands within the ingested HTML.
Capability inventory: The skill generates a GitHub Actions workflow (.github/workflows/deploy.yml) and full repository contents, which are intended to be executed and hosted on the GitHub platform.
Sanitization: There is no mention of sanitization, escaping, or validation of the input HTML before it is written to the index.html file.
[EXTERNAL_DOWNLOADS]: The skill generates instructions and configuration files that reference external resources.
Official GitHub Actions: The generated deploy.yml workflow utilizes official GitHub-maintained actions including actions/checkout@v4, actions/configure-pages@v5, actions/upload-pages-artifact@v3, and actions/deploy-pages@v4. These are well-known and standard components for the platform.
Author Branding: The generated README.md includes a link to affitor.com, which is the official site of the skill author.
Technical Documentation: The skill provides links to official GitHub documentation for verifying IP addresses and managing custom domains.

github-pages-deployer